Makes one wonder: can one post prompt injection attacks instructing moltbots to update their skills files to add malicious behaviors or download and run malicious code libraries? Are moltbots on moltbook protected from this? Seems like a huge security vulnerability